Related: When I was a first-year PhD student, our stats professor described factor analysis as attempts of a shark (or maybe a whale?) to capture as many fish as possible in a school of fish. It would pass through the school multiple times, until it just wasn't worth it anymore.
Exactly! Each pass through the school represents one factor. And the residuals basically reflect the shark having determined it's not worth trying to capture what remains.
Intriguingly cool, but I always wonder whether these types of explanations are helpful? How are students now supposed to think about models with three or four rhs variables? The paper and bees move in the wind (over time) and then?
The bees already raise so many questions in 3D space though. Like it mentally works as long as I imagine that the bees instead behaved like fixed points in space lol.
Comments
"*Mega* Analogie, digga. Schreib das genau."